Understanding the selectivity of small molecule binding is a central question in drug development. It impinges on both determining which desirable targets might respond to the compound as well as which unintended targets might lead to toxicity. Currently these are difficult questions to answer because they require the evaluation of the small molecule's interactions with many proteins, which in turn requires methods for testing many proteins and for detecting and characterizing the interaction. Protein microarrays offer a compelling method for presenting many proteins for testing, but the methods currently available for assessing small molecule binding require modifying the small molecule with a label. This labeling frequently results in changes to the small molecule's binding performance and can be difficult. A technology that enables researchers to address these questions in a high-throughput manner will accelerate the transition of discoveries into the clinic. In this project, we propose to bring two technologies together that will result in a high-throughput system to detect and characterize small molecule-protein interactions using a label-free system that is extremely sensitive, quantitative and provides information on binding kinetics. The two technologies are a label-free multiplexed detection system that is based on nanohole arrays and a protein microarray system, nucleic acid programmable protein array (NAPPA). The nanohole array sensor, technology is a detection method based on a novel photonics device whose key characteristics are: single molecule sensitivity, a very high level of multiplexing, label-free detection, and the generation of kinetic binding data. The NAPPA technology is a method for the miniaturized presentation of thousands of proteins based on in situ expression of protein from cDNAs printed onto a surface. NAPPA addresses the key constraints associated with other protein microarray systems, namely, protein stability and storage, elimination of the need for protein purification, and captures 1000 fold more protein than other approaches. This project also makes use of a library of expression ready cDNA clones for human proteins that is being assembled by the Harvard Institute of Proteomics. These technologies have the potential to produce a valuable tool in lead identification and optimization.